Ask a question
7 cross-paper comparisons
Hand-curated questions with quantitative parameter deltas across papers.
- Macaque vs human RDM psychometric
- Macaque vs human RDM chronometric
- Rat vs human clicks psychometric
- IBL biased prior blocks
A file-first atlas of behavioral neuroscience tasks.
Curated task records, harmonised data, and cross-paper findings on shared
axes — every record is YAML in git, every analysis is reproducible, every
number links back to the dataset that produced it.
The shape of the atlas
Live demo
Refit a curated psychometric, in your browser
The atlas runs Pyodide + scipy locally and fits a 4-parameter logistic to 156 observed points (2,383 trials) from London D. First click downloads ~10 MB of runtime; subsequent fits are instant.
Bars are Wilson 95% CIs from each point's n.
- Species
- human
- Trials
- 2,383 trials
Click Refit live to run a 4-parameter
logistic on this curve in your browser. Parameters appear here once
the fit completes.
Same Python that runs the build-time fits — the page just delegates to your machine. Open finding london-2018-poisson-clicks-dbs.dbs-off for diagnostics, residuals, and the full fit table.
Where to start
Browse by species
91 findings across 4+ species
Curves filtered by organism, on shared signed-evidence axes.
All findings →Browse by source depth
17 reproducible slices
Filter datasets by source-data depth: raw trials, processed trials, figure-traced.
Open catalog →Family coverage
Where the atlas runs deep, where it doesn't
R
raw
P
processed
F
figure
| Family | Findings | Models | Source | Next |
|---|---|---|---|---|
| deep human · mouse · visual | 54 | 117 | source rank 2 | inspect depth |
| deep human · NHP · visual | 28 | 72 | source rank 2 | inspect depth |
| modeled human · rat · auditory | 7 | 29 | source rank 3 | inspect coverage |
| modeled mouse · visual | 2 | 4 | source rank 3 | inspect coverage |
| source-backed mouse · auditory · multisensory · visual | 0 | 0 | source rank 2 | extract findings |
| source-backed mouse · visual | 0 | 0 | source rank 3 | extract findings |
| source-backed mouse · somatosensory | 0 | 0 | source rank 2 | extract findings |
Featured comparison
σ spans 4.11 to 5.04 (1.2× steeper at the steep end)
How do macaque and human random-dot motion psychometrics compare?
Roitman & Shadlen 2002 (two macaques) and Palmer-Huk-Shadlen 2005 (six human observers) both ran the same signed motion coherence 2AFC at processed-trial level. They use the same canonical axis but different species and different report modalities (saccade vs button-press), so a slope difference is a candidate signature of perceptual sensitivity at the species level.
Recently added
As of 2026-07-26 · 91 findings, 18 papers, 12 slices in the atlas.
Recently active · 6 min read
Cross-species random-dot motion
Classic macaque RDM, human reaction-time RDM, and macaque confidence-wagering RDM live on one canonical signed-coherence axis. DDM fits give a single drift parameter to read off across species — and the spread is bigger than you'd expect.
Featured because random dot motion was extended on 2026-04-30.
Psychometric + chronometric + DDM cross-species fits
All stories at /stories.
Across the literature
60 psychometric fits, 18 papers
Slope σ varies 2.8× across species — human fit steepest (3.45), mouse shallowest (9.50).
Median bias (μ) -2.20
Median slope (σ) 4.83
Median 75% threshold 3.52
Σ trials 381,619
Threshold distribution · 11 fits dominant axis only
Each point is one finding's fitted x-at-75%. Click a point to open its detail page. Dashed line marks the across-finding median.
Median σ by species
Filled point is the median σ; light tick marks are the per-finding distribution. Click a median to open the species' findings.
Built 2026-07-26 · ec14ee5